Zero-free regions
From Polymath1Wiki
The table below lists various regions of the [math](t,y,x)[/math] parameter space where [math]H_t(x+iy)[/math] is known to be non-zero. In some cases the parameter
- [math] N := \lfloor \sqrt{\frac{x}{4\pi} + \frac{t}{16}} \rfloor[/math]
is used instead of [math]x[/math]. The mesh evaluation techniques also require rigorous upper bounds on derivatives. In some cases the spacing of the mesh is fixed; in other cases it is adaptive based on the current value of the evaluation and on the derivative bound.
Date | [math]t[/math] | [math]y[/math] | [math]x[/math] | From | Method | Comments |
---|---|---|---|---|---|---|
1950 | [math]t \geq 0[/math] | [math]y \gt \sqrt{\max(1-2t,0)}[/math] | Any | De Bruijn | Theorem 13 of de Bruijn | Proves [math]\Lambda \leq 1/2[/math]. |
2004 | 0 | [math]y\gt0[/math] | [math]0 \leq x \leq 4.95 \times 10^{11}[/math] | Gourdon-Demichel | Numerical verification of RH & Riemann-von Mangoldt formula | Results have not been independently verified |
2009 | [math]t \gt 0[/math] | [math]y \gt 0[/math] | [math]x \geq C(t)[/math] | Ki-Kim-Lee | Theorem 1.3 of Ki-Kim-Lee | [math]C(t)[/math] is not given explicitly. Also they show [math]\Lambda \lt 1/2[/math]. |
2017 | 0 | [math]y\gt0[/math] | [math]0 \leq x \leq 6.1 \times 10^{10}[/math] | Platt | Numerical verification of the Riemann hypothesis | |
Mar 7 2018 | 0.4 | 0.4 | [math]N \geq 2000[/math] ([math]x \geq 5.03 \times 10^7[/math]) | Tao | Analytic lower bounds on [math]A^{eff}+B^{eff} / B^{eff}_0[/math] and analytic upper bounds on error terms | Can be extended to the range [math]0.4 \leq y \leq 0.45[/math] |
Mar 10 2018 | 0.4 | 0.4 | [math]151 \leq N \leq 300[/math] ([math]2.87 \times 10^5 \leq x \leq 1.13 \times 10^6[/math]) | KM | Mesh evaluation of [math]A^{eff}+B^{eff} / B^{eff}_0[/math] and upper bounds on error terms | |
Mar 11 2018 | 0.4 | 0.4 | [math]300 \leq N \leq 2000[/math] ([math]1.13 \times 10^6 \leq x \leq 5.03 \times 10^7[/math]) | KM | Analytic lower bounds on [math]A^{eff}+B^{eff} / B^{eff}_0[/math] and upper bounds on error terms | Should extend to the range [math]0.4 \leq y \leq 0.45[/math] |
Mar 11 2018 | 0.4 | 0.4 | [math]20 \leq N \leq 150[/math] ([math]5026 \leq x \leq 2.87 \times 10^5[/math]) | Rudolph & KM | Mesh evaluation of [math]A^{eff}+B^{eff} / B^{eff}_0[/math] and upper bounds on error terms | |
Mar 11 2018 | 0.4 | 0.4 | [math]11 \leq N \leq 19[/math] ([math]1520 \leq x \leq 5026[/math]) | Rudolph & KM | Mesh evaluation of [math]A^{eff}+B^{eff} / B^{eff}_0[/math] and upper bounds on error terms | |
Mar 22 2018 | 0.4 | 0.4 | [math]x \leq 1000[/math] | Anon/David/KM | Mesh evaluation of [math]H_t[/math] | |
Mar 22 2018 | 0.4 | 0.4 | [math]1000 \leq x \leq 1600[/math] | Rudolph | Mesh evaluation of [math]H_t[/math] | |
Mar 22 2018 | 0.4 | 0.4 | [math]8 \leq N \leq 10[/math] ([math]803 \leq x \leq 1520[/math]) | Rudolph | Mesh evaluation of [math]A^{eff}+B^{eff} / B^{eff}_0[/math] and upper bounds on error terms | |
Mar 23 2018 | 0.4 | 0.4 | [math]20 \leq x \leq 1000[/math] | Anonymous | Mesh evaluation of [math]H_t[/math] | |
Mar 23 2018 | [math]t \gt 0[/math] | [math]y \gt 0[/math] | [math]x \gt \exp(C/t)[/math] | Tao | Analytic bounds on [math]A^{eff}+B^{eff} / B^{eff}_0[/math] and error terms; argument principle | [math]C[/math] is in principle an explicit absolute constant |
Mar 27 2018 | 0.4 | [math]0.4 \leq y \leq 0.45[/math] | [math]7 \leq N \leq 300[/math] ([math]615 \leq x \leq 1.13 \times 10^6[/math]) | KM | Mesh evaluation of [math]A^{eff}+B^{eff} / B^{eff}_0[/math] and upper bounds on error terms; argument principle | |
Mar 27 2018 | 0.4 | [math]0.4 \leq y \leq 0.45[/math] | [math]0 \leq x \leq 1000[/math] | Anonymous | Mesh evaluation of [math]H_t[/math]; argument principle | Completes proof of [math]\Lambda \leq 0.48[/math]! |
Mar 31 2018 | [math]0 \leq t \leq 0.4[/math] | [math]0.4 \leq y \leq 1[/math] | [math]10^6 \leq x \leq 10^6 + 1 [/math] | KM | Mesh evaluation of [math]A^{eff}+B^{eff} / B^{eff}_0[/math] and upper bounds on error terms; argument principle | |
Mar 31 2018 | 0.4 | [math]0.4 \leq y \leq 0.45[/math] | [math]0 \leq x \leq 3000[/math] | Rudolph | Third approach to [math]H_t[/math]; argument principle | |
Apr 6 2018 | [math]0 \leq t \leq 0.2[/math] | [math]0.4 \leq y \leq 1[/math] | [math]5 \times 10^9 \leq x \leq 5 \times 10^9+1[/math] | KM, Rudolph, David, Anonymous | Mesh evaluation of [math]A^{eff}+B^{eff} / B^{eff}_0[/math] and upper bounds on error terms; argument principle | |
Apr 6 2018 | 0.2 | 0.4 | [math]N \geq 3 \times 10^6 (x \geq 1.13 \times 10^{12})[/math] | KM | Analytic lower bounds on [math]A^{eff}+B^{eff} / B^{eff}_0[/math] and upper bounds on error terms | |
Apr 7 2018 | 0.29 | [math]y \geq 0.29[/math] | [math]N \geq 19947[/math] | Anonymous | Triangle inequality bound on [math]A^{eff}+B^{eff} / B^{eff}_0[/math] and upper bounds on error terms | Would in principle show [math]\Lambda \leq 0.33205[/math] if the matching barrier could be established |
Apr 9 2018 | 0.2 | [math]y \geq 0.4[/math] | [math]N \geq 3 \times 10^5[/math] | Tao | Triangle inequality bound on [math]A^{eff}+B^{eff} / B^{eff}_0[/math] and upper bounds on error terms | |
Apr 10 2018 | 0.2 | [math]y \geq 0.4[/math] | [math]4 \times 10^4 \leq N \leq 10^5; 100|N[/math] | KM | Euler2 mollifier and triangle inequality bounds on [math]A^{eff}+B^{eff} / B^{eff}_0[/math] | Error terms not estimated but look well within acceptable limits |
Apr 12 2018 | 0.2 | [math]y \geq 0.4[/math] | [math]4 \times 10^4 \leq N \leq 3 \times 10^5[/math] | Anonymous | Euler2 mollifier and triangle inequality bounds on [math]A^{eff}+B^{eff} / B^{eff}_0[/math] | Error terms not estimated but look well within acceptable limits |
Apr 12 2018 | 0.2 | [math]y \geq 0.4[/math] | [math]19947 \leq N \leq 4 \times 10^4[/math] | Rudolph | Euler3 mollifier and triangle inequality bounds on [math]A^{eff}+B^{eff} / B^{eff}_0[/math] | Error terms not estimated but look well within acceptable limits |